1. Field of the Invention
The present invention relates to an image processing apparatus and an image communicating apparatus.
2. Related Background Art
With the spread of a digital public communication line network (what is called an ISDN), a television/telephone and a television conference system have been put into practical use. The International Consultative Committee for Telephone and Telegraph (CCITT) has determined various rules and examinations with respect to audio, video, and data transmissions in such a digital communication line. For instance, the rule H.320 among those rules specifies the whole system aspect of audio/video services. The rule H.221 specifies a frame structure in the audio/video transmission on the channel of 64 kbps to 1920 kbps and a coding allocation of a BAS (Bit-rate Allocation Signal) which is used for exchange of a terminal capability, designation of a communication mode, or the like.
The rule H.242 specifies a protocol for an exchange of a capability among terminals and a switching of the communication mode by the BAS. As for a moving image, the rule H.261 relates to a moving image signal portion of audio video services at a speed of p.times.64 kbps (p31 to 30) and specifies coding and decoding methods of two video signal formats of CIF (Common Intermedia Format) and QCIF (Quadratic Common Intermedia Format). The QCIF relates to a format such that the CIF is reduced by 1/2 in the horizontal direction and the number of lines is reduced into 1/2.
As mentioned above, according to the recommendation (and draft of the recommendation) by the CCITT, after the physical end-to-end of initial channels and the additional channels and the synchronization establishment of each connected channel (the in-channel) is performed, there are specified procedures for: an exchange sequence of the terminal capability by the BAS in the in-channel; a mode switching sequence by the designation of a communication mode, and the like; and a format of image data which is transmitted among terminals. However, a process to change the terminal capability in accordance with situations in each terminal and which communication mode is used in a range of the exchanged capability are out of the range of the specification. An exchanging method between the television signal of the NTSC system or the like and the CIF or QCIF is also out of the range of the specification. Generally, however, one picture plane of the CIF or QCIF format is converted into one picture plane of the NTSC system.
When one picture plane of the CIF or QCIF format is converted into one picture plane of the NTSC, for example, a roughness of an image is conspicuous as compared with that of QCIF.
On the other hand, when a transmission rate of an image is low, a trackability for the motion is bad. When an image is displayed on the whole screen, therefore, stiffness of the motion of the image is conspicuous.
As a method of coding an image, a block coding method whereby one picture plane is divided into a plurality of blocks each consisting of a predetermined number of pixels and pixel data is coded with respect to each block is widely known. According to the block coding method, what is called a block distortion cannot be avoided. The block distortion strongly appears at a position near the block boundary and is weakened as the position is away from the block boundary.
Hitherto, such a block distortion is eliminated by smoothing the decoded pixel values for the whole screen. For example, the pixel values are smoothed by a simple mean, weighted mean, intermediate value, or the like every pixel with reference to the pixel values around the target pixel. Such a smoothing process is uniformly executed irrespective of the pixel position and also independently of a coding rate of a moving image.
On the other hand, a construction such that the presence or absence of a block distortion is checked with respect to the neighboring pixels of the block boundary and, when there is a block distortion, the smoothing process is executed on only those neighboring pixels has been disclosed in JP-A-3-46482.
In case of a moving image, there is a tendency such that, as a coding rate is small, a block distortion of the reproduction image increases. In the conventional apparatus, since a magnitude of the smoothing process has been set to a predetermined level irrespective of a coding rate, there is a drawback such that in case of a large coding rate such that a block distortion hardly occurs, the smoothing process which is stronger than it is needed is executed, so that the reproduction image becomes blurred or, contrarily, when the coding rate is low, the smoothing process is insufficient.
As mentioned above, the block distortion strongly appears at a position near the block boundary and becomes weak as the position moves away from the block boundary. Therefore, when the smoothing process is uniformly performed on the whole picture plane, the image unnecessarily becomes blurred in a portion away from the block boundary.
In the construction disclosed in JP-A-3-46482, although the block distortion of the neighboring pixels of the block boundary can be eliminated, the block distortion of a region near the block boundary cannot be eliminated.